Porosity Method In The Research And Application Of Liver Ultrasound Image Recognition

Ultrasound imaging technology has been widely used in the medical field In recent years because of its advantages of high sensitivity, easy operation, low cost and without trauma.However, the ultrasound image is mostly exist noise and low resolution.Doctors often diagnose the illness through the naked eyes, This method will easily lead to misdiagnosis,So we need an auxiliary means to help doctors correctly identified ultrasound image and improve the diagnostic accuracy.This paper is based on the theory of the fractal geometry and lacunarity, and had a very good attempt in the ultrasonic liver image analysis. In recent years, lacunarity mean has been widely used in medical image processing and image analysis. This article attempts to use lacunarity mean to analysis the ultrasonic liver images,and to improve the recognition rate of liver cancer. The purpose is use the value of lacunarity to describe the characteristics of ultrasonic liver image texture. The value of lacunarity of ultrasonic liver images are calculated by box-column-average method, and we use support vector machines for classification. Through analysis, we get the following conclusions:(1) In the same scale,the value of lacunarity of ultrasonic normal liver image is significantly less than that of liver cancer.(2) The value of lacunarity by box-column-average method has a strong characterization of ultrasound liver image texture features.The simulation results show that the value of lacunarity can represent some characteristics of the ultrasound liver image, lacunarity is a very good parameter to identify ultrasound liver images,and have a high recognition accuracy,so it can be used as an effective adjunct to diagnosis liver disease.